Interlaboratory comparisons: Lessons learned

Abstract

Interlaboratory comparisons have been widely used in analytical chemistry and radiochemistry as an important part of ongoing quality assurance programs. The 14C community has been no exception in this respect, and in just under 20 years, there have been a number of significant and very extensive interlaboratory trials organized by individual laboratories and the International Atomic Energy Agency (IAEA) to the benefit of the 14C community (both labs and users) (Otlet et al. 1980; ISG 1982; Scott et al. 1990; Rozanski et al. 1992; Scott et al. 1992; Gulliksen and Scott 1995). The comparisons have varied widely in terms of sample type and preparation, but all have had as their primary goal the investigation of the comparability of results produced under possibly quite different laboratory protocols. As techniques have been developed and new labs formed, the reference materials created as a result of the intercomparrisons have presented an opportunity for checking procedures and results. Users have been reassured by the existence of regular comparisons as one sign of the concern that laboratories have to ensure highest quality results, but also confused about how to make use of the results from past exercises in the interpretation of sets of dates from many laboratories. The laboratories have also learned valuable lessons from participation in such studies. These have included identification of systematic offsets and additional sources of variation and in studies which have used realistic samples requiring pretreatment, chemical synthesis and counting, it has been possible to identify the stage at which such problems have arisen and to quantify the relative contributions to the overall variation. In this paper, we will briefly review the comparisons so far, draw some conclusions from their findings, and make proposals for the future organization of intercomparisons.